Advancing Sexual and Reproductive Health and Rights
 
International Perspectives on Sexual and Reproductive Health
Volume 39, Number 2, June 2013
DIGEST

Child Marriage Associated With Suboptimal Reproductive Health

Child marriage is associated with unintended pregnancy, low levels of contraceptive use early in marriage and limited use of maternal health services, according to an analysis of Demographic and Health Survey data from Bangladesh, India, Nepal and Pakistan.1 In at least three of the countries, and sometimes in all four, women who had married before age 18 were more likely than those who had married later to have had an unintended pregnancy (odds ratios, 1.5–3.8), to have had multiple unintended pregnancies (2.3–24.2) and to have not used contraceptives before their first live birth (1.3–2.5). In addition, child marriage was frequently associated with reduced odds of having given birth in a facility and having had skilled personnel at the delivery (0.5–0.8).

Although previous research has linked child marriage with a range of adverse reproductive health outcomes, much of this work has been done in India. To assess whether these associations are widely present in South Asia, researchers analyzed data from the four largest countries in the region that had had a Demographic and Health Survey since 2005—Bangladesh (2007), India (2005–2006), Nepal (2006) and Pakistan (2006–2007). Analyses focused on women aged 20–24 who were married, divorced or widowed; women who did not live with their spouse were excluded. In addition to compiling descriptive statistics, the researchers carried out logistic regression analyses, separately for each country, to identify associations between age at marriage (categorized as 14 or younger, 15–17 or 18 or older) and 10 outcomes; they used Poisson regression to assess associations with an 11th outcome, antenatal visits. Sample sizes for most analyses were about 14,600 for India and 1,500–2,100 for the other three countries; however, analyses of rapid repeat childbirth (giving birth within 24 months of a previous delivery) were restricted to women with at least two children and used much smaller samples (6,400 for India and 700–1,000 for the remaining countries). All regression models controlled for women’s age, residence (urban vs. rural), household wealth, educational attainment, religion, geographic area and age gap between spouses. In addition, analyses of contraceptive use adjusted for number of living sons, and those of use of maternal health care adjusted for birth order.

Although the legal age at marriage is 18 in Bangladesh, India and Nepal, early marriage was common in all three countries, especially in Bangladesh, where 39% of respondents had married during middle adolescence (ages 15–17) and 38% during early adolescence (14 or younger). The proportions who had married during middle adolescence were even higher in India (42%) and Nepal (50%), although the prevalence of marriage during early adolescence was much lower in these countries (17% and 12%, respectively). Child marriage was least prevalent in Pakistan, where the age of consent for marriage is only 16; some 36% of respondents had married during middle adolescence, and 14% during early adolescence.

For two fertility measures, no evidence of detrimental outcomes emerged among women who had married as children. Although at least one in six women in each country had given birth within the first year of marriage, the odds did not differ according to the timing of marriage. Rapid repeat childbirth, which was reported by 33–66% of women, was negatively associated with child marriage in India and Nepal (odds ratios, 0.6–0.7), and not associated with age at marriage in the other countries.

Some 14–24% of women in each country had had an unintended pregnancy, and 2–4% had had more than one such pregnancy. Both outcomes were positively associated with child marriage. In every country but Bangladesh, the odds of having had an unintended pregnancy were elevated among women who had married at age 14 or younger (odds ratios, 1.6–2.9) or at ages 15–17 (1.5–3.8), and in all four countries, women in one or both early marriage groups were more likely than women who had married later to have had multiple unintended pregnancies (odds ratios, 2.3–24.2). Marriage during early or middle adolescence was consistently and positively associated with having had a pregnancy that ended in miscarriage, stillbirth or abortion (the three outcomes were grouped together, as per the source data); the associations were weakest in India (1.4–1.6) and strongest in Bangladesh (3.4–4.5).

The researchers examined three measures of contraceptive use. In every country but Pakistan (for which data were unavailable), women who had married before age 15 were less likely than those who had married at age 18 or older to have used contraceptives before their first live birth (odds ratios, 1.5–2.2); findings were similar among women who had married at ages 15–17 in India and Bangladesh (1.3–2.5). Perhaps because women who had married early had attained their desired family size at a young age, child marriage was often positively associated with current contraceptive use. In India and Bangladesh, women in one or both child marriage groups were more likely than those who had married as adults to be using a modern method (odds ratios, 1.5–2.6), and child marriage was positively associated with having been sterilized in India (5.4–10.4) and Nepal (4.7–9.1), the only countries where the number of women who had undergone the procedure was sufficient to permit analysis.

Analyses of maternal health care use were restricted to women’s most recent birth. The number of antenatal visits they had had generally did not differ by age at marriage, although Bangladeshi women who had married before age 15 and Nepalese women who had married at ages 15–17 reported 8–15% fewer visits than did their counterparts who had married at age 18 or older.* Child marriage was frequently, but not consistently, associated with reduced odds of having delivered at a health care facility and having had skilled delivery personnel; in every country, the odds of one or both outcomes were lower among women in at least one of the child marriage groups than among women who had married as adults (odds ratios, 0.5–0.8).

The study’s limitations, the authors note, include its reliance on self-reported, cross- sectional data; the small samples in some analyses; and the grouping of stillbirths and spontaneous and induced abortions. Nonetheless, the findings suggest that in South Asia, “child marriage is significantly associated with poor fertility outcomes, lower contraceptive use early in the marriage, and inadequate maternal health care use,” and underscore the potential value of interventions that target women who married early. Unfortunately, designing and implementing such interventions can be a challenge, the researchers add, because women who marry early often leave school, do not work outside the home and have restricted mobility, “which makes them a hard-to-reach population for reproductive health programs.”—P. Doskoch

REFERENCE

1. Godha D, Hotchkiss DR and Gage AJ, Association between child marriage and reproductive health outcomes and service utilization: a multi-country study from South Asia, Journal of Adolescent Health, 2013, 52(5):552–558.

*Results for this outcome were reported as coefficients rather than as odds ratios because they were from Poisson regression models.